EP0184971B1

EP0184971B1 - Molten carbonate fuel cell with improved electrolyte storage

Info

Publication number: EP0184971B1
Application number: EP85630212A
Authority: EP
Inventors: Stephen John Szymanski; Harold Russel Kunz
Original assignee: United Technologies Corp
Current assignee: RTX Corp
Priority date: 1984-12-10
Filing date: 1985-12-03
Publication date: 1989-08-02
Also published as: CA1249330A; EP0184971A2; JPS61140072A; EP0184971A3; JPH0821395B2; DE3572089D1; ZA859227B; US4596751A; BR8506032A

Description

The field of art to which this invention pertains is molten carbonate fuel cells and more particularly to molten carbonate fuel cell with improved anodes, cathodes and matrices.
Fuel cells which use alkali-metal carbonates as the electrolyte are well known in the art and are generally referred to as molten carbonate fuel cells since the electrolyte is liquid at typical operating temperatures in the range of 550°C-750°C. The electrolyte is usually mixed with an inert particulate or fibrous material which remains solid during cell operations, maintains the space between the cathode and anode portions of the cell, and prevents the mixing of the two reactants. The electrolyte is in direct contact with the electrodes where the three phase reactions (gas-electrolyte-electrode) take place. Hydrogen is consumed in the anode area producing water, carbon dioxide, and electrons. The electrons flow to the cathode through an external circuit producing the desired current flow. At the anode there must be ready entry for the reactant gas, ready exit for the chemical reaction products and ready exit for the product electrons.
Molten carbonate cells, like other fuel cell types rely on liquid capillary forces of electrolyte contained in the "matrix" to seal and isolate reactant gases. Thus it is essential that the matrix should be filled with electrolyte, at least as much electrolyte as is necessary to fill the matrix is introduced into the fuel cell prior to its being put into operation for the first time. However, during fuel cell operation electrolyte is gradually lost through evaporation, creepage, reaction, and removal in product gas streams. Thus to preserve the seal function and to maintain ionic continuity within the cell, electrolyte should be stored in molten carbonate fuel cells in excess of that needed for initial cell operation.
Electrochemical cells which can accommodate changes in electrolyte volume have been disclosed. Note commonly assigned US-A-4,035,551, US-A-4,062,322 and US―A―4,038,463. For instance, US―A―4,035,551 discloses an electrolyte reservoir for a fuel cell, including molten carbonate cells, comprising an electrode substrate with a pore distribution that in conjunction with the electrolyte matrix provides a source of electrolyte from the electrode to the matrix. These inventions significantly increase the cell's ability to respond to changes in the volume of electrolyte thus increasing the cell's performance.
However, in order to reach the 40,000 hours life necessary for commercial success, there is a continual search in the art for stable molten carbonate cells having longer life at high levels of performance.
A principal object of the present invention is a long life metal carbonate fuel cell containing an anode electrode, a cathode electrode in contact with and separated by an electrolyte matrix where all three components contribute to electrolyte storage and an accommodation of electrolyte losses during the operation of the fuel cell by specific physical characteristics in line with claim 1.
A preferred feature of the invention is set forth in dependent claim 2.
In accordance with the present invention greater fuel cell stability and performance are attained without adding or subtracting materials, parts, or components. These results are achieved by designing components based on their respective sensitivity to electrolyte loss so that for instance, the matrix, being most sensitive to electrolyte loss, retains the highest electrolyte level as electrolyte is lost. The parameters affected include mean pore size, pore size distribution and porosity. This invention is a significant advancement towards reaching a stable long-life fuel cell that is necessary for commercial success.
The foregoing and other objects, features and advantages of the present invention will become more apparent from the following description of preferred embodiments and accompanying drawings.

Figure 1 plots normalized volume as a function of the pore diameter for fuel cell components;
Figure 2 plots component void volume as a function of pore diameter for fuel cell components having specific characteristics;
Figure 3 plots component void volume as a function of pore diameter for fuel cell components, where one component has different specific characteristics than in Figure 2; and
Figure 4 translates the data of Figure 2 and Figure 3 into performance by plotting performance loss as a function of fill.

The cell components (anode, cathode, and matrix) can be made from any material known in the art to be suitable for molten carbonate fuel cell components, as the improved performance and stability of this discovery are based not on the materials but the form of the components. However, the base metal particles are preferably nickel, cobalt, and mixtures thereof. Electrodes could also include conductive dispersoid particles or stabilizing agent particles which preferably contain aluminum, chromium or zirconium.
The particle size and particle size distribution are selected to impart the cell components with certain characteristics pertaining to % porosity, mean pore size, pore size distribution, etc. These characteristics (described below) provide the cell components with the form that results in improved performance and stability. The correlation between particle size and the enumerated characteristics can be readily understood by those skilled in the art.
The anode has a shape conventional to the art, however its form is such that it has several - characteristics detailed below. For instance it has a mean pore diameter less than the cathode's, about 10 pm, and preferably less than about 4 µm. The anode should have a porosity sufficient to contain enough electrolyte to provide for the electrolyte lost through evaporation, creepage and reaction. This is typically about 45% to about 65% porosity and preferably about 55% porosity. Percent porosity is defined as the pore volume divided by the total component volume (including the pore volume). The anode should also have a ratio of maximum pore diameter at 90% electrolyte fill to maximum pore diameter at 10% electrolyte fill less than about three. Maximum pore diameter refers to the largest pore size that contains electrolyte at a certain level of electrolyte fill. The term % fill refers to the component void volume filled with electrolyte as a percentage of the total component void volume. Preferably, 90% of the pore volumes have a pore diameter less than about 7 pm and 10% of the pore volumes have a pore diameter less than about 2.5 µm. In addition best results are achieved when the initial electrolyte volume in the anode is larger than that of the cathode and matrix.
The cathode has a shape conventional to the art; however, again its form is such that it has several characteristics described below. It should have a mean pore diameter larger than that of the anode's and preferably a mean pore diameter above about 10 pm. The cathode should be above about 0.05 cm in thickness and preferably about 0.07 cm to about 0.10 cm in thickness. Also the cathode's pore size distribution is such that, when the anode electrolyte fill changes from 90% to 10% fill, the cathode electrolyte content should change from about 40% to about 25% fill. Preferably, 60% of the pore volumes have a pore diameter less than about 14 µm and 30% of the pore volumes have a pore diameter less than about 4 µm. The cathode should also have a porosity above about 78%. This is equivalent to about 60% after experiencing cell environment (the cell in operational mode). Generally, cathode porosities change when placed in typical cell operating conditions, i.e., hot oxidizing gases (649°C). In addition, the cathode should have about 20% of its pore volume in pores less than about 0.5 um in diameter resulting in a surface area above about 1 m²/g.
The matrix has a shape conventional to the art but should have a pore size distribution such that greater than 98% of the matrix is filled with electrolyte when the anode is 10% filled with electrolyte. Preferably, 98% of the pore volumes have a pore diameter less than about 0.8 pm. It is also preferred that the matrix has a mean pore diameter of about 0.15 pm.
Although each of the above-described components independently will result in improved cell performance, even greater cell performance can be achieved by combining these components to make improved cells.
Thus a preferred embodiment of this invention is a fuel cell with an anode and a matrix both having the characteristics described above. Another preferred embodiment of this invention is a fuel cell with a cathode and a matrix both having the characteristics described above. Still another preferred embodiment of this invention is a fuel cell with an anode and a cathode both having the characteristics described above. An especially preferred embodiment of this invention is a fuel cell with an anode, cathode and matrix that have the characteristics described above.
Fuel cell components of this type may be made by techniques conventional in this art. The stabilizing agent powder and metal powder should be uniformly mixed, the relative concentration and sizes of the metal and dispersoid powders must naturally be such as to produce, upon further processing, the desired microstructural characteristics in the finished body. However, the porosity and pore size distribution in the finished component is influenced not only by the concentrations and sizes of metal and dispersoid particles and sintering conditions but also by the degree of compaction experienced in forming the compact. The compact may be formed by any of the various well-known techniques, including uni- or multidirectional die pressing, isostatic pressing, powder rolling, extruding, and roll bar molding.
The sintering of the formed powdered metal articles is preferably accomplished in a sintering furnace having an inert or reducing atmosphere, usually hydrogen. The sintering temperature depends on the type of metal particles utilized for both the base metal and the dispersoid particles. Generally sintering is carried out at a temperature which is approximately 95% of the melting point of the base metal. The sintered article is then preferably compacted by any conventional means to form an article having the desired degree of porosity. The article can then be subjected to an annealing treatment if required. The exact temperature and duration of annealing depends on the materials used to form the porous metal article.

Example

A cathode was made using the following process. Nickel powder, available from Inco Co. as #255 was sifted through a #100 mesh screen having a sieve opening of 149 pm avialable from Tyler Co. A mold mask was filled with the sifted nickel powder. The mold powder was compacted by rolling to impart the desired porosity i.e. the more rolling the less porosity. The mold mask was then transferred onto a belt running through a furnace for sintering at a temperature greater than 650°C for 3045 minutes and temperatures in the range of 954-1027°C for 8-12 minutes. The sintering atmosphere was 20% hydrogen and 80% nitrogen and a minimum oven belt speed of 7.62 cm minute was maintained. The sintering conditions also affect the electrodes final characteristics. The resultant structure was removed from the furnace, cooled and rolled to a thickness of 0.100±0.005 cm. Subsequent to experiencing cell conditions, i.e. hot oxidizing gases (0₂, CO₂, H ₂0 at 649°C) in which the nickel cathode is transformed into nickel oxide, the cathode had the approximate characteristics of cathode A in Table I; thickness 0.075 cm, porosity 60%, mean pore diameter 10.5 µm 37% initial fill and 25% final fill. Anode and matrix material according to the present invention can be similarly made, with appropriate adjustment of the above process including the starting materials, the degree of compaction and the sintering conditions.
The components of this disclosure are used in molten carbonate fuel cells. Generally, the operating temperatures of a molten carbonate fuel cell range from about 550°C to about 750°C. At the cathode, oxygen and carbon dioxide react to form carbonate via the overall reaction:
At the anode, hydrogen in the fuel gas reacts with carbonate ions from the electrolyte to form water and carbon dioxide:
It should be noted that while this disclosure is specifically directed to molten carbonate fuel cells, it would be within the purview of one skilled in the art to make components that have specific characteristics relative to electrolyte loss sensitivity for other fuel cell types, i.e. phosphoric acid fuel cells.
These cell components provide improved performance and stability because they have been optimally designed based on their respective sensitivity to electrolyte loss. While not wanting to be held to any theory, it is believed that electrolyte sensitivity is determined by two competing concerns. One factor is that the electrolyte should be minimized so that the diffusion of reactant gas into the electrode does not cause a concentration polarization. However, if there is too little electrolyte, there will be an increased ohmic loss which results from the difficulty of ions to migrate from one electrode to another. Although there is an optimum level of electrolyte volume for each component, a different working range exists in which each component will function. In order of increasing sensitivity to electrolyte loss the components are anode, cathode and matrix. This discovery ensures that electrolyte will be lost from the components in proportion to their respective sensitivity to electrolyte loss. It achieves this by optimal design of, for instance, pore design so that through capillary action electrolyte will be lost from the component with the largest pore size first.
A clearer understanding of this discovery and the component's relative sensitivities to electrolyte loss may be had by reference to the accompanying figures. In Figure 1 normalized volume (Y₁) is plotted as a function of the pore diameter (X_i) for a matrix (11), anode (12), cathode A (13) and cathode B (14).
In Figure 2 the component void volume (Y₂) in cubic centimeters (cm³) is plotted as a function of pore diameter (X₂) in microns (pm). The void volume distributions of the anode curve (21), cathode curve (22), and matrix curve (23) when totalled equal the total cell volume distribution curve (24). Such a diagram was obtained by taking the individual pore distributions in Figure 1 and multiplying the normalized volume by the corresponding void volume of each component (anode, cathode A and matrix). Thus void volume is defined as (Frontal Area, sq. cm) (Thickness, cm) (Porosity, %)/100.
To illustrate the discovery suppose that the cell package initially contains 0.115 cm³ (25 in Fig. 2) of molten carbonate electrolyte per sq. cm. frontal area for a corresponding anode fill of 90% (26 in Fig. 2). All pores with a diameter less than or equal to 6.8 Ilm will then be flooded. With time, electrolyte is lost due to evaporation, creepage and chemical reaction. Assume that after a given period of time only 0.042 cm³ (27 in Fig. 2) of molten carbonate electrolyte per sq. cm. frontal area for a corresponding anode fill of 10% (28 in Fig. 2). The component characteristics and initial and final states of the electrodes and matrix are detailed in Table I.
Therefore, of the 0.073 cm³ lost, 0.067 cm³ or 92% was from the anode and 0.006 cm³ or 8% was from the cathode. The matrix remains full. Since the anode is more tolerant to electrolyte volume changes, the anode is used for storage.
Figure 3 demonstrates the impact of small changes in component design on electrolyte loss. The figure follows the same concept as Figure 2, however there were two minor changes made to the cathode, the pore size was increased from 10.5 µm to 16 µm and the thickness was increased from 0.076 to 0.089 cm. Thus component void volume cm³ (Y₃ in Fig. 3) is plotted as a function of pore diameter µm (X₃ in Fig. 3). Again the normalized volume of the individual pore distributions in Figure 1 were multiplied by the corresponding void volume of each component. However, the changed cathode is represented by curve 14 in Figure 1, and curve 31 in the void volume distribution diagram Figure 3. The void volume distributions of the anode (32) and matrix (33) are also represented. Again the volumes of each component for a given pore diameter have been totalled to give the cell volume curve (34). The component characteristics and initial and final fill states of the electrodes and matrix are detailed in Table II.
Using the same initial (35) and final (36) cell electrolyte contents (0.115 cm³ and 0.042 cm³ respectively) as in the previous case, 0.073 cm³ electrolyte was again lost. However this time 0.063 cm³ or 87% was lost from the anode and 0.010 cm³ or 13% was lost from the cathode while the matrix remains full. In comparison to the previous case on a relative basis more electrolyte was lost from the cathode and less from the anode. This is the reverse of their respective sensitivities to electrolyte loss.
Figure 4 illustrates performance for cathodes similar to those described above. Cathode performance loss in millivolts (mv) at 160 milliamps/square centimeter (mA/cm²) (Y₄) is plotted as a function of estimated percent cathode fill (X₄). Cathode A (41) experienced a performance loss of only 32 to 35 mV as the cell and cathode fill levels varied from 76% to 28% and 37% to 25% respectively. By contrast cathode B (42) experienced a performance loss of 300 to 160 mV as the cell and cathode fill levels varied from 73% to 26% and 30% to 11 % respectively. This severe penalty is a result of the characteristics of cathode B, specifically, the loss of volume in pores with diameters less than 0.5 pm and the higher mean pore size.
This discovery provides a significant advancement in fuel cell technology by improving the overall performance of the fuel cell by defining component design parameters. These parameters include pore size, pore-size distribution and porosity. The components maintain cell performance and stability as electrolyte is lost due to reaction, leakage and evaporation during normal cell operations. Thus the anode, cathode, and matrix are designed based on their respective sensitivity to electrolyte loss so that for instance, the matrix, being most sensitive to electrolyte loss retains the highest electrolyte level as electrolyte is lost.
This discovery increases the cell's ability to respond to changes in the volume of electrolyte without adding or subtracting materials, parts or components. It marks another step towards the long-life operation necessary for the commercial success of fuel cells.

Claims

1. A fuel cell including an electrolyte matrix containing an alkali metal carbonate which is molten during operation of the fuel cell, the matrix having a mean pore diameter of about 0.15 µm, an anode electrode having a mean pore diameter of less than about 4 pm in contact with the electrolyte matrix and a cathode electrode having a mean pore diameter above about 10 um spaced apart from the anode electrode and in contact with the electrolyte matrix wherein the anode has a porosity sufficient to contain enough electrolyte to provide for the electrolyte lost during normal operation of the cell and has a ratio of maximum pore diameter at 90 percent electrolyte fill to maximum pore diameter at 10 percent electrolyte fill of less than about 3 such that more than 98 percent of the matrix is filled with electrolyte when the anode is 10 percent filled with electrolyte and wherein the cathode has a pore size distribution such that when the anode electrolyte content changes from 90 to 10 percent fill, the cathode electrolyte content changes from about 40 to about 25 percent fill, the cathode having a porosity greater than about 78 and has about 20 percent of its pore volumes in pores less than about 0.5 µm in diameter resulting in a surface area defining the pores above about 1 m²fg.

2. Fuel cell according to claim 1, characterized in that the cathode is 0.075 to 0.100 cm thick.